Distinguishing High-Performing Health Systems Using a Composite of Publicly Reported Measures of Ambulatory Care

Ann Intern Med. 2020 Nov 17;173(10):791-798. doi: 10.7326/M20-0718. Epub 2020 Sep 22.


Background: Payers and policymakers are rewarding high-performing health care providers on the basis of summaries of overall quality performance, and the methods they use for measuring performance are increasingly important.

Objective: To develop and evaluate a measure that ranks health care systems by ambulatory care quality.

Design: Systems were ranked using a composite model that summarizes individual measures of quality, accounts for their correlation, and does not require health care systems to report every measure. The composite measure's suitability was evaluated by examining whether it captured the quality indicated by component measures (validity), whether differences in rank between health care systems were larger than statistical noise (reliability), and whether year-to-year changes in rank were small (stability).

Setting: California and Minnesota, 2014 to 2016.

Participants: 55 health care systems.

Measurements: Publicly reported measures of ambulatory care quality.

Results: The composite measure was valid in that it was broadly representative of the component measures and was not dominated by any single measure. The measure was reliable because the ranks for 93% of California systems and 80% of Minnesota systems were unlikely to be more than 2 places lower or higher. The measure was stable because fewer than half of systems changed ranks by more than 2 ranks from year to year.

Limitation: The analysis is limited to available measures of ambulatory care quality and includes only 2 states.

Conclusion: This composite measure uses publicly reported data to produce valid, reliable, and stable ranks of ambulatory care quality for health care systems in Minnesota and California, and this approach could be used in other applications.

Primary funding source: Agency for Healthcare Research and Quality.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Ambulatory Care / standards*
  • Ambulatory Care / statistics & numerical data
  • California
  • Delivery of Health Care / statistics & numerical data*
  • Minnesota
  • Models, Statistical
  • Quality Indicators, Health Care* / statistics & numerical data
  • Reproducibility of Results